README.txt
1.87 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
This directory (lre07) contains example recipes for the 2007 NIST
Language Evaluation. The subdirectory v1 demonstrates the standard
LID system, which is an I-Vector based recipe using full covariance
GMM-UBM and logistic regression model. The subdirectory v2 demonstrates
the LID system using a time delay deep neural network based UBM
which is used to replace the GMM-UBM of v1. The DNN is trained using
about 1800 hours of the English portion of Fisher.
The following LDC corpora are used during training:
SRE 2008 training set: LDC2011S05
CALLFRIEND Vietnamese: LDC96S60
CALLFRIEND Tamil: LDC96S59
CALLFRIEND Japanese: LDC96S53
CALLFRIEND Hindi: LDC96S52
CALLFRIEND German: LDC96S51
CALLFRIEND Farsi: LDC96S50
CALLFRIEND French: LDC96S48
CALLFRIEND Standard Arabic: LDC96S49
CALLFRIEND Korean: LDC96S54
CALLFRIEND Mainland Chinese Mandarin: LDC96S55
CALLFRIEND Taiwan Chinese Mandarin: LDC96S56
CALLFRIEND Caribbean Spanish: LDC96S57
CALLFRIEND Non-Caribbean Spanish: LDC96S58
LRE 1996: LDC2006S31
LRE 2003: LDC2006S31
LRE 2005: LDC2008S05
LRE 2007 Training Set: LDC2009S05
LRE 2009: LDC2014S06
Note that some of the corpora, e.g., SRE 2008 and the LREs used for
training contain multiple languages. Because of this, it isn't
necessarily vital that all of the corpora are present in your system.
The NIST 2007 Language Evaluation (LDC2009S04) is used for testing.
This list will be updated as scripts for system development and testing
(which will require additional data sources) are created.